396 
H. E. Jordan 
nucleus. Tlicir niimber is so large, however, that it is impossible to 
decide definitely concerning tlieir growth, nor is it possible to prove 
that some may not arise de novo in tlie nucleus, though this seems 
improbable. 
Düring earlier stages tlie primary nucleolus enlarges in spite of its 
contribution of secondary nucleoli, but in later stages it bas usually en- 
tirely disappeared. The final stages in its disintegration are marked by 
Fig. 3. 
Still later stage in the growth period showing the dispersion of the nncleolar hnds (accessory nucleoli). 
The latter (hlack) lie scattered, more densely peripherally, throoghout the pale purple nucleus. Fixation 
has produced a wide contraction artifact. The cytoplasm is beginning to change centxally from the 
granulo^reticular to the alreolar type and stains a brownish pink. X dOO. 
a shrivelled appearance of its surface and a loss of basophile staining 
reaction of its central portion. As such it may persist to the last, its 
final dissolution being accomplished by fragmentation. Occasionally, 
however, the primary nucleolus appears to remain spheric and chromatic 
to the last, its ultimate remnant becoming indistinguishable from the 
secondary nucleoli. 
The important point of difference between Cribreüa and Echinaster 
in respect to the origin of secondary (accessory) nucleoli is that in the 
former they arise from the primary (chief) nucleolus by a process of ex- 
trusion of chromatic spherules, in the latter by fragmentation. In the 
